Copyright © 2021 by the author(s). Published here under license by the Resilience Alliance. Choudhury, M.-U.-I., C. E. Haque, A. Nishat, and S. Byrne. 2021. Social learning for building community resilience to cyclones: role of indigenous and local knowledge, power, and institutions in coastal Bangladesh. Ecology and Society 26(1):5. https://doi.org/10.5751/ ES-12107-260105

Research Social learning for building community resilience to cyclones: role of indigenous and local knowledge, power, and institutions in coastal Bangladesh Mahed-Ul-Islam Choudhury 1, C. Emdad Haque 1, Ainun Nishat 2 and Sean Byrne 3

ABSTRACT. Despite wide recognition of the role of social learning in building community resilience, few studies have thus far analyzed how the power–knowledge–institution matrix shapes social learning processes that in turn foster resilience outcomes. Drawing insights from the biopolitical lens of resilience, we take a critical stance on programmatic interventions for community resilience and social learning, arguing that local knowledge, beliefs, practices, and social memory (SM) are crucial elements in social learning processes for building community resilience to shocks and stresses. In addition, we explore how technologies shape social learning processes and build or strengthen community resilience. Our research, conducted in cyclone-prone coastal zones of Bangladesh, adopts a transformative interpretive framework (TIF) and a community-based participatory approach to empirical investigation. Findings of our research provide evidence that formal institutions frequently exclude indigenous and local knowledge (ILK) from social learning processes, and often subjugate communities to notions of resilience, as defined by nonlocals, that perceive people as subjects of institutional power and objects of scientific knowledge, rather than as active agents. We further found that local communities are able to obtain early warnings of cyclones through ILK of environmental phenomena, such as changing water temperature and animal behavior. Despite an abundance of ILK regarding past cyclones, the 2007 Cyclone Sidr was a surprising event to many and caused considerable loss of life and property. Much of this unpreparedness stemmed from an overall SM deficit—a key to translating knowledge into action. We recommend strengthening efforts to bridge scientific–technical knowledge and ILK to ensure effective social-learning-led resilience outcomes are achieved. Key Words: Bangladesh; community resilience; cyclone; disasters; indigenous knowledge; local knowledge; power; social learning; social memory

INTRODUCTION learning and building resilience require an understanding of these The enhancement of community resilience to emerging risks, such dynamics (Cannon and Müller-Mahn 2010, Gaillard 2010, Carr as climate-change-induced extreme weather events, is an issue that 2019). In this regard, some critical social theorists appropriately has received a great deal of attention in both academic and policy assert that resilience is essentially progressive and political as it domains (Cutter et al. 2008, Haque et al. 2018). Recognizing that envisions people as active agents who have control over their own resilience is often socially and politically differentiated, for the destiny, rather than as passive subjects and victims (Grove 2013b, purpose of the present study, we broadly defined “resilience” as 2014b, Evans and Reid 2014, Barrios 2016). Support for this the (inherent) ability of a community to withstand external shocks human agency perspective has also been registered in some and disruptions (Folke el al. 2010, Berkes and Ross 2013, Faulkner applied community resilience scholarship, e.g., focusing on et al. 2018). This fundamentally depends upon the community identifying, building, and nurturing local strengths (Brown and members’ capacity and opportunity for learning from crises and Westaway 2011, Berkes and Ross 2013, Faulkner et al. 2018). In combining different forms of knowledge in order to effectively this paper, we posit that an inter- and transdisciplinary prepare for and respond to future crises (i.e., social learning) engagement with the work of these critical social theorists and (Adger et al. 2005, Folke et al. 2003, Berkes 2007, Pahl-Wostl 2009). applied, pragmatically oriented scholars helps to better With this basic relationship in mind, we investigated risk reduction understand the process of resilience on the ground. and community resilience to nature-triggered extreme events As a mechanism for sharing and developing common (NTEE) and associated disasters through a social learning lens, understandings, social learning for building resilience often and posit that indigenous and local knowledge (ILK) is vital to the involves external deliberative processes (Pahl-Wostl 2006, social learning process and for enhancing resilience to NTEE. Armitage et al. 2008), and several authors have raised concerns Numerous recent studies have rightly elaborated the problematic regarding who defines resilience—and for whom (MacKinnon nature of normative framings of resilience (Cote and Nightingale and Derickson 2013, Cretney 2014). They assert that if external 2011, Christensen and Krogman 2012, Fabinyi et al. 2014, Brown deliberation (e.g., by formal institutions) does not meaningfully 2016) and the fact that resilience has frequently been found to be include local voices, knowledge, and memory, these processes are socially (e.g., gender), politically (e.g., power), and culturally (e.g., likely to subjugate people to the power of institutions and local knowledge) differentiated (Pelling 2011, Grove 2014a, b, scientific–technical forms of knowledge (Adger et al. 2001, Bulley Brown 2016, Jordan 2018). As the normative usage of resilience 2013, Grove 2014b). Here, unpacking the relational matrix of may obscure its relationship with vulnerability, facilitating social power, knowledge, and the institutional context is essential for

1Natural Resources Institute, University of Manitoba, Canada, 2Centre for Climate Change and Environmental Research, BRAC University, Bangladesh, 3Arthur V. Mauro Institute of Peace and Justice, University of Manitoba, Canada Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

understanding the role of social learning in building community different stakeholders participate in a collective learning platform resilience, which is often ignored in the social learning discourse. to share their experiences and opinions, learn from each other, and come to a common understanding of the issues that Understanding tensions and differentials in power and knowledge contribute to adaptive resilience (Cutter et al. 2008, Johannessen is central to comprehending and explaining the “everyday forms and Hahn 2013, Baird et al. 2014). Social interactions and of resilience” that are discursively produced and reproduced networks are thus central to the social learning process. However, through social learning processes (Pelling 2011, Boyd et al. 2014, it is also essential to critically examine such a normative Brown 2016). Recognizing the plurality of approaches to conceptualization of social learning by investigating how power– knowledge, our intent is to explore diverse meanings and knowledge plays out within institutional contexts that in turn possibilities for alternative forms of resilience (Grove 2013a, 2018, shape social learning processes and resilience outcomes (cf. Pugh 2013, Bonilla 2020). In this study, by taking a critical stance Armitage et al. 2008, Reed et al. 2010, Boyd et al. 2014, Ensor on programmatic interventions for community resilience and and Harvey 2015). social learning, we argue that local knowledge, beliefs, practices, Despite a wide array of literature underscoring the importance and social memory (SM) are critical elements of social learning and utility of ILK for building community resilience to shocks processes for building community resilience to NTEE, such as and stresses (Thomalla and Larsen 2010, Gómez-Baggethun et cyclones and associated storm surges. As well, we assert that it is al. 2012, 2013, Kelman et al. 2012), the incorporation of local important to examine how these elements are being applied in voice in the social learning process that is embedded in the idea power relations and how technologies play out in such social of ILK has thus far been poorly studied (Briggs and Sharp 2004). learning processes. Social learning must build on local strengths (e.g., ILK and SM) In our investigation, we consider ILK as a point of convergence as these represent vital inner capacities or capital of a community between social learning and community resilience scholarship, (Magis 2010). Hooli (2016) extends this notion by asserting that and further link these fields to that of disaster resilience. We achieving resilience among the poorest requires the incorporation document and analyze the effects of various dimensions of ILK of their knowledge and learning. Studies of social learning in reducing risks and building resilience to NTEE, the interplay processes that focus on the implications of the inclusion and/or between ILK and SM in shaping resilience, and the role of exclusion of local knowledge for building community resilience community-based institutions in replenishing SM and in the are nonetheless scant in the literature. social learning process. We then present a critical stance on social It is difficult to operationalize social learning in the case of learning and resilience to interrogate the role of external formal episodic events like NTEEs. In natural disaster research literature, institutions in this process. In so doing, we aim to demonstrate, learning and knowledge are often used interchangeably (Pfister with empirical evidence, how formal institutions exclude ILK, 2009), but social learning can be both a “process” through which and thus how social learning activities that promote more knowledge is acquired and produced or an “outcome” in and of inclusive ILK can improve outcomes and facilitate alternative itself (Pahl-Wostl 2006, Armitage et al. 2008, Pfister 2009). For practices of resilience. the purposes of this study, we consider ILK (including SM of CONCEPTUAL CONSIDERATIONS disasters) as a surrogate of social learning. Indigenous and local knowledge is fundamentally social in character as it is situated Indigenous and Local Knowledge, Social Learning, and within wider social entities or communities of practice. It is gained Community Resilience to Nature-Triggered Extreme Events through a process of continuous accumulation from empirical The capacity of actors to learn and to combine different forms observation and trial and error, transmitted from one generation of knowledge (i.e., adaptability) is a prerequisite for building to the next, and embedded within local institutions and practices resilience (Folke et al. 2003, 2010). Various forms of learning and (Dekens 2007, Berkes 2018, Trogrlić et al. 2019). its outcomes are conceptualized in resilience literature, such as incremental, episodic, transformative, and social learning Indigenous and local knowledge reflects communities’ inner (Holling 2004, Berkes 2007, Gunderson 2010, Pelling 2011). We strengths, which have the potential to improve preparedness as posit that, from a resilience perspective, learning implies system- well as to reduce risk and enhance community resilience to NTEE oriented learning, i.e., social learning (Berkes 2007, Pahl-Wostl (Berkes 2007, Kelman et al. 2012). It is a source of community 2009), which can be incremental (single- and double-loop) or resilience and adaptive capacity (Boillat and Berkes 2013, Gómez- transformative (triple-loop). We consider social learning a useful Baggethun et al. 2013). Supporting and enhancing local strengths lens to interrogate the connection between ILK and community is also a key to community-based risk reduction (Thomalla and resilience and have adopted Reed et al.’s (2010: 6) view of social Larsen 2010, Choudhury et al. 2019). However, ILK is often learning as “a change in understanding that goes beyond the narrowly defined in disaster risk reduction (DRR) and resilience individual to become situated within wider social units or literature, where the focus is primarily on acquired knowledge, communities of practice through social interactions between ignoring social institutions, power relations, beliefs, practices, actors within social networks.” memory, and worldviews. In this respect, it is useful to draw insights from Berkes’ (2018) framework for traditional ecological Social learning has diverse meanings, applications, connotations, knowledge, which defines a “knowledge–practice–belief and acquisition processes, and often involves individual-, complex” consisting of “a cumulative body of knowledge, network-, and system-centric approaches (Rodela 2011, 2013). In practice, and belief, evolving by adaptive processes and handed the climate-induced disaster risk and resilience context, it can be down through generations by cultural transmissions, about the considered a process (deliberative and/or spontaneous) wherein relationship of living beings (including humans) with one another and with their environment” (Berkes 2018: 8). Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Analogous to Berkes’ (2018) four interrelated levels of analysis, framework with strong reciprocal relationships. Although all we offer a framework to investigate the function of ILK in DRR levels of analysis are important, for the purposes of this study, and resilience (Fig. 1). Indigenous and local knowledge is the first we focus on the ILK and the role of social institutions in reducing level of analysis and is at the core of this framework. Whereas risk and building and/or strengthening community resilience. Berkes (2018) developed his framework in the context of natural resources management and was largely concerned with the Social Memory and Community Resilience to Nature-Triggered ecological aspects of ILK (e.g., species identification and Extreme Events behavior), our work focuses on disaster risk and community Social memory is an “arena in which captured experience with resilience. We expand this level by drawing insights from Acharya change and successful adaptations, embedded in a deeper level of and Prakash’s (2019) six dimensions of ILK (including ecological) values, is actualized through community debate and decision that are relevant to DRR and resilience: (i) ecological (nonhuman making processes into appropriate strategies for dealing with behavior); (ii) phenomenological (anticipation of a probable ongoing change” (Folke et al. 2005: 453, also see Hewer and Kut disaster based on memory); (iii) sea/riverine; (iv) meteorological; 2010, Beilin and Wilkinson 2015). However, SM has diverse (v) celestial; and (vi) official (information received from external meanings and connotations (Olick 2016). We consider SM institutions and electronic and print media). These dimensions of specifically in relation to NTEE and disaster shocks and its role ILK often interact to provide early warning before the onset of in DRR and community resilience building. disasters, which is likely to facilitate actions to prepare, reduce Social memory is essential to a community's capacity to respond risks, and enhance resilience (Trogrlić et al. 2019). The to shocks and it is a source of renewal and self-organization phenomenological dimension includes SM, which has special (Berkes 2007); it is argued that communities with higher SM tend significance for DRR and resilience (see section on SM below). to be more resilient (Folke et al. 2005). Human responses and The official dimension indicates that people often proactively resilience strategies in the face of disaster are shaped by the integrate their own knowledge and experience with external presence or absence of SM, as it is required for translating knowledge and modern technologies (Hilhorst et al. 2015). knowledge into action (i.e., the link between the first and second levels of our framework). However, SM has two major drawbacks: Fig. 1. Nested levels of analysis in ILK for DRR and resilience it tends to fade away with the passage of time, and it may provide [after Berkes 2018]. a false sense of confidence. Longer intervals between events tend to make SM less reliable. Intangible SM in the form of narratives and oral history is likely to fade away quickly if not renewed by recent disaster experience or vigorous institutional efforts. It is likely to be held by people who directly experienced past events, typically elders (Berkes and Folke 2002), and in the absence of a renewal mechanism, “only half of the population remembers the most intensive and extensive natural processes after ten years, and only a tenth after forty years” (Komac 2009: 206). Furthermore, because SM is based on past experience, it can provide overoptimistic and inaccurate expectations regarding the likely extremes of climate- change-induced events, and lead to surprises when future events exceed all prior experiences. Institutions, Social Learning, and Resilience Building In the second level of analysis, people perform local resilience The third level in Fig. 1 comprises social institutions—the formal practices based on existing knowledge and memory. At this level, and informal rules, regulations, and social norms governing a social learning often emerges from people’s direct encounters with community (Ostrom 2008)—that can play a decisive role in risks and catastrophic events. The third level of analysis involves nurturing, replenishing, and sharing SM both horizontally (actor formal and informal social institutions, which profoundly affect to actor) and vertically (generation to generation). Socially performance regarding DRR and community resilience (see embedded informal institutions play a critical role in generating section on institutions below). The fourth level of analysis and disseminating social learning and in the “memorialization” concerns worldviews, belief, and values, which shape people’s process (Tidball et al. 2010, Rumbach and Foley 2014). Formal attitudes and responses to NTEE. Risk reduction and resilience institutions are increasingly taking responsibility for reducing to NTEE are a function of worldviews, belief, and values (Fig. disaster risks and building resilience to NTEE by creating 1), and these may be altered by exposure to NTEE and disaster collective learning platforms and promoting new scientific risks (Oliver-Smith 2002). knowledge and ideas. In light of this trend, a vital question that has emerged is: do formal institutions take local voices, learning, Two aspects of this framework are noteworthy here: ILK as a ILK, and SM into consideration in the social learning process, or “knowledge–practice–belief complex” is not static but rather a is the social learning process mostly top-down? Taking a critical dynamic process that involves trial and error and the integration stance on the role of formal institutions in social learning and of new ideas and knowledge; and the four interrelated levels are resilience-building processes, we examine to what extent not hierarchical, but rather are linked under a single nested institutions shape the pathways of resilience (Wilson 2014). Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Critical scholars argue that “resilience” is a contested concept community-based disaster resilience programming is alleged to with multiple meanings, interpretations, and political and ethical become the technology of biopower. The intimate operation of implications (Grove 2013a, 2018, Pugh 2013, Bonilla 2020). power to promote so-called experts’ notions of resilience, which Addressing the question of “who” defines “resilience” and for Agrawal (2005) describes as “intimate institutions,” is also “whom,” MacKinnon and Derickson (2013) have noted that illustrated by this perspective. nonlocals (e.g., external donors and development agencies) often Thus, a biopolitical reading of resilience is concerned with the define community resilience, whereas the voices of the techniques that produce knowledge to naturalize risks and communities themselves are silenced. Understanding tensions uncertainties and to intervene in the lives of communities to help and differentials in power and knowledge in such processes is them cope and adapt to such changes (Chandler 2018, 2019). central to comprehending and explaining the alternative Evans and Reid (2013) explain that, in such analyses, resilience meanings of resilience (Brown 2016, Pelling 2011, Boyd et al. operates through the ontology of vulnerability, which Chandler 2014). In discussing the relationship between power and (2019) termed “ontopolitics.” Therefore, following biopolitical knowledge, Foucault ([1980], 2005) explains that through rationalities, the intent of resilience technologies is to make exercises of institutional power and the exclusionary processes subjects resilient so that they have the capacity to adapt and to involved, one form of knowledge becomes “truth” and others are exploit situations of uncertainty in the face of a multiplicity of rejected. Such processes may take place at local levels where locally threats (Evans and Reid 2013, Hill and Larner 2017). The relevant knowledge may be excluded by the power of prevailing biopolitical interpretation considers agents as active and having institutions (Agrawal 2005). control over their own lives, and criticizes traditional However, for Foucault, power is not always oppressive and interpretations of resilience for considering (resilient) subjects as constraining but can also be productive, giving rise to new forms powerless and lacking agency and for ascribing any vulnerabilities of interest, desire, capacity, and behavior. Such a to disasters to deficiencies in the subjects’ abilities rather than conceptualization of power views individuals as active agents flaws in existing disaster management systems (Bockstael 2017). rather than passive subjects (Foucault [1978], 2005). The In the social learning process, ILK should therefore be considered Foucauldian notion asserts that “the truth of resilience is not ‘out in its own right for encouraging the local and indigenous peoples there’, objectively waiting to be discovered,” but rather that the to pursue their own notion of resilience. As Barrios (2016: 35) conceptualization and definitions of resilience are constructed argues, “[d]efinitions of resilience, recovery, and “rebuilding through power relations. Thus, “critical research on resilience has better,” must not only be polyvocal, but must also foreground the shown that resilience initiatives create subjects with particular voices of people and communities who directly bear the brunt of kinds of desires and capacities” (Anonymous reviewer, 3 July disasters,” otherwise resilience-building processes may instead 2020, personal communication). Foucault argues that “where there enhance vulnerability. In addition, in building resilience, the goal is power there is resistance;” people are not merely victims of should be to provide resources to people and enable them to make hegemonic norms, but rather strategically cope with the their own choices, rather than implementing interventionist discursively produced and reproduced social domination (Cleaver strategies (Kevin and Jonathan 2015, Evans and Reid 2014). 2007, Grove 2013a). Therefore, resistance is not always an overtly manifested act, but rather an expression of hidden, STUDY AREA AND METHODOLOGY unconventional, and strategic positions (Brown 2016). Foucault’s We conducted our study in the southern coastal zone of biopolitical analysis of power is also relevant in explaining the Bangladesh (Barguna district), which is ranked by shift of the regimes of power, knowledge, and technologies Intergovernmental Panel on Climate Change (IPCC) models as beyond the state apparatus (e.g., toward NGOs) (see Grove among the most vulnerable areas (IPCC 2014). Over the last half 2014b). Biopolitics signals a “problem space” (i.e., hazards and century, recurrent severe tropical cyclones, originating in the Bay disaster risks are governmental concerns) and a topological of Bengal, have made landfall into this region. Following the analysis of power “that examines how existing techniques and devastating human toll (approximately half a million fatalities) technologies of power are redeployed and recombined in diverse of Cyclone Bhola (11 November 1970), a 3–5 m elevated earthen assemblies of biopolitical government” (Collier 2009: 79). dike system was developed to protect the coastal region, In recent years, a redeployment of power and authority has taken particularly from cyclone-associated storm surges. In recent years, place both globally and locally, wherein state roles and Barguna district was most severely affected by Cyclone Sidr in responsibilities for disaster management have been delegated to 2007, which claimed 1292 lives, injured 16,310 people, and local authorities according to what is commonly termed the destroyed a great deal of property and infrastructure “subsidiarity principle” (Melo Zurita et. al. 2015, 2018). Such (Government of Bangladesh (GoB) 2008). Cyclones Aila (2009), delegation has frequently been performed by the state through Mahasen (2013), and Komen (2015) have subsequently devastated partnering with NGOs and other civil society organizations the district. We carried out our field investigation from August (Choudhury et al. 2019). This approach assumes that local 2018 to January 2019 in two Upazilas (subdistricts) of Barguna institutions, by virtue of being embedded in the community and district, namely Amtali and Taltali, covering six villages—two building social capital and trust, are better equipped to deal with from Amatali (Baliatali and Gupkhali) and four from Taltali local shocks and stresses (Melo Zurita et al. 2018). Notably, these (Nidrarchar, Idupara, Tatulbaria, and Nolbonia) (Fig. 2). subsidiarity principle-based approaches often sidestep fundamental We adopted a transformative interpretive framework (TIF) questions regarding how people are governed through (Mertens 2007, Creswell 2013) in studying social learning decentralization processes and what effects such processes have processes for building community resilience to coastal cyclones upon peoples’ vulnerabilities. From a biopolitical perspective, and associated risks. The TIF framework acknowledges that Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

knowledge and its production within a society are value laden— participant, following our protocol as approved by the University that is, they reflect power asymmetry and shape social of Manitoba’s (Canada) Joint-Faculty Research Ethics Board. relationships. Therefore, the intent of knowledge production must Considering the scope and objectives, the selection of respondents be to change the lives of people and the institutions they live in in the empirical investigation was made purposefully. We first (Mertens 2010, Creswell 2014). Community voices and conducted 50 semistructured interviews with community participation are critical to the production of knowledge that is members who had firsthand experience with Cyclone Sidr. This likely to facilitate transformative change and coping with talk-based method helped capture community members’ diverse oppression and subjugation (Mertens 2012, Thiessen and Byrne narratives on memory, experience, and learning from past major 2017). Parallel to resilience thinking, this paradigm focuses on the NTEE and associated disasters. We interviewed adult and elderly “strengths” within marginalized and disadvantaged communities male and female members (30 males and 20 females). Interviews rather than their “deficiencies” (Mertens 2009, Brown and ranged from 30 to 90 minutes. The interviews included exploring: Westaway 2011). experiences of past major disasters; coping and adaptation strategies; what was learned from disaster experience; actions and Fig. 2. Location map of the study area. roles performed by local formal institutions; application of learning to later events; and how early warnings (EWs) were formulated based on ILK prior to the disaster event. In the second stage, we conducted 11 focus group discussion (FGD) sessions involving diverse occupational groups from six villages. Focus group discussions enabled capturing people’s narratives on learning and experience with NTEE and disasters. We conducted six FGDs with males (two with farmers, three with fishermen, and one with a mixed group) and five with females. These diverse FGDs were organized in order to capture diverse dimensions of ILK, more recurrent features, and outliers. Community members interact with surrounding environments differently for livelihood due to their varied occupation and gender positions. Therefore, differential experiences and exposures to risks allow people to generate ILK on multiple and varied dimensions. For example, fishers are more capable of observing and documenting the maritime and riverine dimensions of ILK than farmers, who concentrate more on terrestrial dimensions. In the third stage, we collected six oral histories (four males and two females) from elders (aged more than 65 years) who had experience of catastrophic cyclones before Cyclone Sidr (2007). This method made space for a “voice” or a “picture” of the past from the words and memories of the respondents. The main purpose of this investigation was to unpack the function of memory and elders’ knowledge in building and maintaining resilience. The duration of these conversations ranged from 75 to 105 minutes. In the final stage, we conducted five key informant interviews (KIIs) with representatives from community-based local institutions (e.g., local press club) to understand their roles in the Our empirical investigation followed a community-based memorializing and social learning process. The local press clubs participatory approach. This is a transformative change approach in Amtali and Barguna shared images and news reports that to research that aims to unshackle people from oppressive- highlighted their roles in this process. hegemonic power structures (Jacobson and Rugeley 2007), and in turn, facilitate change in the lives of participants as well as “the Because formal institutions from outside the community institutions in which they live and work” (Creswell 2007: 21). Our profoundly shape the social learning and resilience building primary data collection involved techniques drawn from the process, we conducted six KIIs with NGO personnel to participatory rural appraisal (PRA) toolbox, with a four-tier understand how they carry out DRR and resilience projects at study design (Choudhury and Haque 2016). Before commencing the local level. We also collected project documents to examine fieldwork, the first author recruited a local field assistant who how external institutions view resilience and DRR. The first worked as a gatekeeper and translator of local dialects. Prior to author participated as a “participant observer” in three social collecting the data, several informal visits to the communities with learning sessions organized by local NGOs to disseminate the field assistant helped build rapport with community members. knowledge on DRR and resilience. Sample questions included in Appropriate verbal or written consent was obtained from each the observation protocol were: who participates in the platform Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Table 1. Dimensions of LK and EW to reduce the risk of coastal cyclones and storm surges

Dimensions of LK LK as EW signs and signals i) Ecological (nonhuman behavior) a. Ants stay together and climb to higher places b. Leafs of Mantha tree get curled c. Cows start starvation d. Seabirds and ducks start moving toward the shore** e. Churi (Eupleurogrammus muticus) and Loittia fish (Harpadon nehereus) start moving and jumping quickly in the sea*** f. Mosquitoes stick into the body of cows and goats g. Flying insects are less visible at night. They seem to be in a rush. h. Guyalla birds fly out to sea i. Livestock such as ducks and hens become reluctant to enter their shelters j. Small black birds start flying over the sea ii) Phenomenological (feeling and a. Elders sense an impending disaster based on past experience and memory ** anticipation of a probable disaster from b. Unbearably hot and humid weather for several days past memory) iii) Sea/riverine (behavior and observation a. Change in the color of seawater ** of river and sea) b. Warm seawater ** c. Water whirls and more vapors in the sea surface air d. Unusual patterns of water movement and flow in the sea e. Sea becomes darker f. Elevated water levels in the adjacent rivers iv) Meteorological (related to wind a. Wind from the southeast or east (locally known as Eshan Kun or pubal wind) brings water surges *** movement, cloud, and temperature) b. Hot and humid weather brings bad weather c. Wind from the west brings rain** d. Wind from the southwest is followed by recession of surge water *** e. When juba (high spring tide) combines with an easterly wind v) celestial (condition of sky and moon) a. Lunar day and month and associated high and low tides *** b. Lightning in the north/ northeast means strong wind and floods (this gives 10–12 h of warning) ** c. Lightning in the southwest means only rain d. If there is no thunder and lightning and if weather becomes silent (gombir), then there is the possibility of storm e. Lightning during the onset of storm reduces its intensity vi) Official (i.e., information received by a. Warning signals from radio, TV, and mobile phones ** external institutions, electronic and print b. Calls from family members and relatives media) c. Warning signals received from CPP (Cyclone Preparedness Program) volunteers *** Most widely reported by community people ** Moderately reported Source: Field data, 2018

(i.e., gender and age); do NGO-facilitated platforms take ILK whereas wind from the southwest or west causes surges to recede and memory into account; do they disseminate and incorporate (Table 1, iv (d)). Community members repeatedly reported that only scientific–technical knowledge in local programming or are prior to the landfall of Cyclone Sidr, an easterly wind (pubal batas) other forms of knowledge included? was blowing for about 24 h, and continued until the water surge receded: “During Sidr, pubal batas was blowing, as soon as wind FINDINGS AND ANALYSIS started moving to west then water started to retreat. We start to Dimensions of Indigenous and Local Knowledge, Early Warning, worry when pubal batas blows because it increases water [as storm Disaster Risk Reduction, and Resilience surge].” People also reported that a few days before the landfall From the point of view of social learning, DRR, and resilience, of Cyclone Sidr, the weather remained hot and humid (Table 1, we focused on two aspects of ILK: triangulation of ILK, and the ii (b)). Often people draw inferences combining multiple possession and use of ILK within the community. Regarding the dimensions. For example, fishermen observed a change in the first aspect, we categorized local early warnings (EWs) for DRR behavior of Churi (Eupleurogrammus muticus) and Loittia and resilience under the six dimensions of ILK (Table 1). Often (Harpadon nehereus) fish, accompanied by a change in color and the role of ILK in DRR and disaster resilience is temperature of the ocean surface (Table 1), whereupon they overromanticized, and its validity is not examined. We tried to attempted to return to shore ahead of the impending storm. triangulate some of the ILK-informed EWs against two other People in coastal Bangladesh often triangulate their own knowledge types: in relation to the onset of a particular cyclone observations with cyclone EWs from official sources (e.g., disaster event (i.e., Sidr), and against established scientific Bangladesh Meteorological Department) and take measures to explanations of pertinent phenomena. The most common and reduce risk from the potential cyclone’s impact. One elder widely reported local cyclone EW feature is the direction and explained that Signal Numbers One, Two, and Three from official rotation of blowing wind. Wind from the southeast or east (pubal sources are “normal” but Signals above Number Seven are batas) pushes water surges toward the locality (Table 1, iv (a)), considered “dangerous” locally. Explaining the use of multiple sources of EW, one FGD member stated: Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Table 2. Possession of various dimensions of LK by different groups of people

Ecological Phenomenological Sea/Riverine Meteorological Celestial Official Occupational Fishers *** ** *** *** *** *** group Farmers * * * *** ** * Others * -- -- * ** ** Gender Male *** *** *** *** *** *** Female -- ** -- *** ** * Age group (in Below 50 ** -- *** *** ** *** years) Above 50 ** *** ** *** *** * *** most widely reported ** moderately reported * less frequently reported

If the sky gets cloudy and pubal batas [easterly wind] The possession and use of different dimensions of ILK within blows, we assume that the weather will become bad soon. the community is helpful in understanding the interaction Then we check radio announcements [broadcasts] and between social learning, DRR, and community resilience. We use SMS services in mobile for weather information. If examined how the structural location of individuals shapes the we hear Signal Number One, we start sailing toward the possession of different dimensions of ILK by documenting the shore thinking that it may become worse soon. structural locations of community members according to occupation, gender, and age. These groups of people were not Established scientific evidence matches some of the EW features mutually exclusive and often overlapped (Table 2). For example, and signs most widely reported and applied by the local people. fishers and farmers are all male and belong to both adult and Our primary intention here is not to validate ILK with scientific elderly age groups. Our findings indicate that specific groups hold knowledge or to nullify; rather, it is to triangulate ILK from certain dimensions of ILK more than others (Table 2). For various pertinent sources. example, ecological dimensions were most widely reported by The first sign used by fishers to predict impending cyclones is a fishers (who are mostly male), whereas the major determinant for significant increase in sea surface temperature (SST). In the reporting the phenomenological dimension (i.e., SM) was age and scientific literature, the relationship between cyclogenesis and SST experience (both male and female). Persons above 50 years old is well established, with cyclonic activity generally occurring when have experienced past disasters firsthand and acquired knowledge the SST exceeds 26°C (Henderson-Sellers et al. 1998, Trenberth through observation and intergenerational knowledge transfer. 2005). As the SST increases, the intensity of cyclonic storms is Fishers’ direct interactions with the sea and rivers made them the likely to increase as such storms are fueled by water vapor (Khan major holders of the sea/riverine ILK dimension. Meteorological et al., 2000, Kossin 2017). During the tropical storm Nargis and celestial dimensions are not influenced significantly by (2008), which eventually made its landfall in Myanmar causing structural locations. Fishers, males, and adults (below 50 years) the death of more than 130,000 people, the SST of the Bay of tend to look for EWs from official sources more than females. Bengal was recorded to be over 30°C (Maneesha et al. 2012). Such variation of ILK within the community in terms of structural locations and other determinants validates the Scientific evidence also supports fishers’ observation that the importance of social learning processes for community resilience color of the sea changes prior to cyclones, as increasing water being system-wide (i.e., community) learning (Berkes 2007, Pahl- temperature triggers an increase in phytoplankton, whose Wostl 2009). pigment in turn absorbs sunlight and further raises the SST (Hernandez et al. 2017, Zhao and Wang 2018). It must be noted Interplay between Indigenous and Local Knowledge and Social that this relationship may not always be linear due to other Memory in Shaping Resilience intervening variables at different temporal and spatial scales A key question that arises from the preceding section is: why do (Dunstan et al. 2018). community people suffer from loss and damage despite possessing The second major sign used by local communities to gauge the rich ILK? We posit that responses and resilience strategies in the onset and intensity of cyclonic storms is wind direction. face of NTEE and disasters are largely shaped by the state of SM, Specifically, an easterly wind (pubal batash) brings storm surges which is a necessary condition for translating knowledge into and tidal flooding—a relationship confirmed by meteorological action (Fig. 1). Here, we elaborate our points through an research (Wicks and Atkinson 2017). A third major sign observed examination of the interplay between SM and ILK in two by fishers is the behavior of coastal fish species (e.g., temporal phases: pre-Cyclone Sidr memory and its impact during Eupleurogrammus muticus and Harpadon nehereus) in response the cyclone itself, and the current state of Cyclone Sidr memory to rising SST. Secor et al. (2019) and Spampinato et al. (2014) and its role in later events. observed such a change in the behavior of fish species during PreSidr memory, indigenous and local knowledge, and resilience tropical cyclones, although they did not specifically link this We have documented four interrelated factors that made Cyclone behavioral change to SST. Sidr a surprise for most of the communities, despite their receiving Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

EWs from multiple sources, including government agencies (e.g., Another source of negative SM is feelings of mistrust toward volunteers from the Cyclone Preparedness Program (CPP), official EW announcements (Fig. 3). Before Sidr, people often relatives, and neighbors). The first factor is the absence of SM of received false disaster warnings, typically from government cyclonic disasters. The last major cyclone disaster before Sidr in agencies. For instance, there was a cautionary tsunami warning 2007 was Cyclone Bhola in 1970. The 37-year gap left most along the coastal areas in September 2007 (2 months before Sidr) community members with no fresh memories of the previous due to the Sumatra earthquakes and Indian Ocean tsunami, disaster, leading many to not take EWs obtained from ILK or forcing many communities to evacuate. No tsunami arrived, and government agencies seriously. Focus group discussion members subsequent warnings about Cyclone Sidr were widely ignored. A explained why they were not able to apply their knowledge during respondent noted: Sidr: We used to ignore early warnings because nothing No one here [present for FGD] witnessed any event like happened after reception of cyclone warnings; therefore, that before. Probably only one had such an experience. we did not trust early warning announcements. We did Then how could we know that? ... Sidr started in the late not believe that Sidr would actually happen. afternoon; we did not realize that can actually happen. In addition to the absence of positive SM and the presence of Elders used to say about floods, but we did not experience negative SM, Sidr came as a surprise to many communities anything like that before... because the celestial dimension of ILK was unable to predict the The second factor is the nonlinearity of cyclonic events. Before danger of a storm surge. Locals in these areas use lunar phases Sidr, some people had memories of storms (locally called dabar) to predict the tides, with the term Juba used to denote high water and their associated ocean-water surges, but few had experienced levels (high spring tides) and Dala to denote low water levels (low severe cyclones and the associated storm-surge-induced flooding spring tides) (see Fig. S.1 in Supplementary Material). As Cyclone (cf. Brammer 1990). Monsoon flooding normally occurs slowly Sidr struck during Dala and low tide, locals did not anticipate a enough to allow time to prepare for its arrival, but the onset of a storm surge over 15 feet in height was possible. cyclone-induced storm surge is much faster than rainwater or In contrast, the presence of positive SM in combination with ILK riparian flooding. People in the studied communities had can make a significant difference in terms of DRR and resilience. experienced floods in 1972 and in 1988, but nothing similar to the We found that, in coastal Bangladesh, community elders, for scale of Cyclone Sidr. As one participant stated: whom Cyclone Sidr was not a novel event, held primarily positive In case of past floods, it took 2 to 3 days for flood water SM. An elder lamented that, “I told everyone that if pubal batash to rise to 6 to 7 feet, or a maximum 10 feet. During Sidr, [easterly wind] does not weaken within 24 h, there will be a storm water came and receded within a half an hour; it washed surge and flood. No one believed me.” In some cases, however, away everything within just blink of eyes ... we did not the memories of the elders, combined with EWs from ILK and experience anything like Sidr before ... we could not government agencies facilitated the undertaking of DRR actions imagine the magnitude of the event. before the onset of the cyclone. Positive SM is likely to trigger responses necessary for reducing Drawing on their SM and ILK, elderly members could foresee risks and building resilience, whereas negative SM can have the impending risks and potential disaster and help younger opposite effect, making people reluctant to take action. One community members avoid property damage and loss of life. One source of negative SM is people’s false confidence in structural young adult respondent (age 35) stated that, “[m]y father warned flood-prevention measures, such as embankments (Fig. 3). As no us about a probable storm surge and flood; he asked us to take community members had previously seen floodwater shelter and store some rice in a safer place.” encroachment inside the dikes, they assumed such dikes would be able to withstand Cyclone Sidr’s surge waters. Memory and learning from Sidr for later events After Cyclone Sidr in 2007, EWs were issued prior to Cyclones Aila (2009), Mahasen (2013), Komen (2015), Roanu (2016), and Fig. 3. Temporal dimension of SM and resilience implications. Mora (2017). During those events, locals took official government EWs more seriously and more willingly undertook measures, such as evacuation, to reduce risks. Some of the survivors of Cyclone Sidr recounted how narrowly they escaped injury or death, the actions they took to cope with the immediate impact of the cyclone as it made landfall (e.g., climbing big trees, grabbing big plastic containers, and rushing to the embankment for safety), and their ignorance of the potential severity of the storm prior to its arrival. Many community members have since embraced the lessons learned from Cyclone Sidr (Table 3). For example, a large proportion now immediately rushes to cyclone shelters upon receipt of government EW. However, since there have been no cyclones on the scale of Sidr since 2007, memory of catastrophic disaster experience has been gradually eroding, and people are becoming more reluctant to evacuate to shelters. As one key Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Table 3. Learning, coping and adaptation practices in study communities

Dimensions Specific learning Coping Taking shelter either on the embankment or in cyclone shelters or preparing to evacuate entirely; Returning from the sea upon receiving EW; Taking measures to save property and valuable goods before evacuating; Preparing to take shelter in a neighbor's house with a stronger structure; Keeping locally available materials (e.g., bamboo trees and big plastic containers) ready to cope with coastal flooding; Adaptation Raising house platforms; Moving houses away from the riverbank or seashore and close to or inside the embankment; Learning that children, (especially pregnant) women, the elderly, and people with disabilities are more vulnerable and require priority aid; Building houses with a stronger structure;

informant observed that the number of people who immediately built in a local cemetery, and local newspapers publish feature flee to shelters was much higher in the years immediately following articles on Cyclone Sidr every year (Fig. 4). Sidr compared with the time of the study (2018): We received early warning during Sidr, but we could not Fig. 4. Observing Sidr Day on 15 November by the Bangladesh realize the severity as we did not experience anything like Red Crescent Society, Barguna Unit (4.a), and Barguna Press that in recent past. We take shelter now if we receive the Club (4.b). Photo courtesy of: Barguna Press Club. early warning. During Mahasen and Aila, most people left their houses for safety. After Sidr, people began to take warnings more seriously, which they did not do before. People did not understand and used to ignore early warnings before... Community-Based Institutions, Memorialization, and Social Learning This section focuses on the role of socially embedded practices and other local institutions (e.g., local media) in facilitating social learning and the memorialization process (the third level in Fig. 1). One important mechanism for social learning embedded in Formal Institutions, Denial of Indigenous and Local Knowledge, rural Bangladeshi culture is the adda (hanging around), an and Processes of Making Resilient Subjects informal platform for sharing personal experiences. The location In Bangladesh, there has been a major institutional shift regarding of adda varies for men and women, with men typically gathering disaster management in recent decades, with many disaster in village marketplaces to socialize over a cup of tea and women management initiatives being decentralized and handed over to gathering in a courtyard. Information gathered is then shared with local institutions, especially through partnering with NGOs the rest of the participants’ families. (Choudhury et al. 2019, Haque and Uddin 2013). This process is In these informal sharing platforms, community members shared driven by an increased focus on local resilience in the national their memories of survival during Cyclone Sidr and drew lessons policy discourse, which is reflected in the National Plan for from each other’s experiences. One key lesson was how to Disaster Management (2016–2020): Building Resilience for recognize existing risks and take them more seriously. Stories Sustainable Human Development (Ministry of Disaster related to survival and death matched with elders’ advice and Management and Relief (MoDMR) 2017). In addition, knowledge (e.g., not to panic and rush or try to evacuate during international donor agencies have implemented numerous large the onset of the storm surge). During our field research, Cyclone projects for building resilience, such as the National Resilience Titli (2018) was forming in the Bay of Bengal. The resulting Program (US$2.25 M) run by the United Nations Development atmospheric depression generated continuous rainfall, during Program. Local-level disaster management institutions, such as which many community members gathered in local tea-shops to NGOs, have been implementing community-based disaster hold adda and discuss their previous experiences with severe resilience projects, where the notion of resilience has been weather. Adda thus functions as an important and effective predefined by policy makers and external donor agencies (Fig. mechanism for replenishing SM. 5a, b). Local institutions, such as the Bangladesh Red Crescent Society Local-level NGOs implementing disaster resilience projects (local unit), local press clubs, and other community-based consider NTEE and associated disasters to be a technical, organizations play a critical role in social learning and the financial, and biological problem. With a predetermined memorialization process. For example, the local press club in the framework, their fundamental intention in reducing risks and study area organizes a memorial event to remember Cyclone Sidr building disaster resilience is to train people in strategies and on every 15 November; they organize rallies, show videos of Sidr, techniques that can help save “biological lives.” Interventions by and hold group discussions. A memorial to the disaster was also NGOs here reflect the government’s biopolitical agenda: the Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

national government frequently presents itself as a success story, Development Centre (CODEC), and JAGO NARI) formulated stating that it succeeded in significantly reducing NTEE-related community-based organizations (CBOs) comprised primarily of fatalities through institutional interventions (Paul 2009). As young and adult men and women. Through these CBOs, monthly Marchezini (2015) argues, state agencies often create a false sense courtyard meetings with women were organized where they of optimism via these claims, and distract attention from the discussed disaster preparedness along with other social issues, actual needs of locals. such as health, hygiene, domestic violence, and child marriage. With male CBO members, NGOs provided training, with Fig. 5. Billboard of the project on disaster resilience (5.a); a supporting technical manuals, on evacuation, rescue, and first aid, community training session flag numbers and EW signals (5.b). and conducted scenario exercises such as mock drilling. A respondent stated that: “They [the NGOs] came after Sidr and taught us how to tackle disasters. ... They told us that we have to keep our eyes open for any surprising extreme events that may stem from natural or other forces. ... If early warning signal moves up to number 10, as we become aware [prepare], we can reduce loss and damage. ... We learned from Climate-Resilient Ecosystems and Livelihoods (CREL) Project.” Second, people feel motivated to participate in NGO projects and In the study area, in line with biopolitical rationalities, local social learning platforms as they expect some material gain from NGOs have created social learning platforms where ILK, their participation. As most NGOs work with so-called “hard- memory, beliefs, and practices are largely ignored. As well, people core” poor, disaster resilience projects are often paired with health are treated as objects of scientific and expert knowledge and and livelihood-building initiatives (e.g., providing sewing subjected to externally defined notions of resilience, as subjects machines, cash, or materials to install a tube well or toilet). We assumed to be vulnerable and deficient in their ability to cope observed evidence of some people’s strategic participation in with shocks associated with coastal cyclones. Resilience to coastal NGO-led learning platforms primarily for material gain as way cyclones therefore implies local people understanding and to address poverty-related suffering and supplement income, internalizing the meaning of the official EW signs and signals and which can be regarded as the “hidden script” of such responding to them in a preordained manner, e.g., by evacuating participation. Community members who did not receive any to a cyclone shelter. Nongovernmental organizations inform material incentives from NGOs tended to withdraw from the people through official EW systems (Fig. 5b), which were learning platforms. One woman explained: “[t]hey gave [sewing originally established during the British colonial period for machine] to others. I went to their office for training for 3 years seaports (Roy 2012). The Bangladesh Meteorological but did not get one ... this is why I do not take part any more.” A Department continues to generate EWs and associated signals for male responded similarly: “I took training with Caritas [an NGO] sea and river ports, which are provided to volunteers of the for 3 years, but I did not get any benefit from them. When they Cyclone Preparedness Program to be disseminated to local donated goats and other things, they did not give them to me. ... communities via the Department of Disaster Management. At the end I withdrew myself.” Nongovernmental organizations also train people on preparedness, risk reduction, and postdisaster recovery strategies Third, we observed that the predesigned frameworks of most (Fig. 5 a, b). NGOs did not allow elders a space in the social learning process, leading to their knowledge and memory being systematically We observed three interrelated factors that facilitate this top- excluded. An elderly respondent stated: “I am old, why would down learning mechanism and subject-making process. First, the they call me? ... They want young, who can walk or travel to go absence of preSidr cyclone memories and traumatic memories of for training in Amtali and Patuakhali, and Barisal [at distant the Cyclone Sidr disaster have created an opportunity (i.e., locations]. They recruit mostly young people, and do not consider “problem space”) for formal institutions to intervene in people’s the elderly at all.” Because the institutions hold power and draw lives. Before Sidr, there were no or merely nominal intervention authority from technical-expert knowledge, exclusion of ILK programs or projects on DRR and resilience. The preceding from their social learning processes is common, whereas, in the discussion highlights that young and adult people had no cyclone local context, ILK is the more valid form of knowledge. Through disaster memory prior to Sidr, which contributed to their these types of practices, local organizations were seen to function traumatic experience of Sidr. Education programs by NGOs on with a preset structure to include and exclude various community DRR and resilience appeared to be attractive for these local people. community members in the absence of positive SM. Explicit exclusion of ILK has been registered not only from the In the study area, all of the six surveyed NGOs have adopted a social learning processes, but also in local DRR and management community-based and participatory method to implement decision making. As part of decentralization and localization of projects for DRR and enhancing resilience, as defined by the disaster management, the local governmental institutions at the formal institutions, to cyclones and storm surges so that people Union Parishad (UP) level—the lowest level of the administrative can cope and adapt to surprises, shocks, and catastrophes. hierarchy—are responsible for planning, including evacuation Initially, NGOs (e.g., Nazrul Smriti Sangsad (NSS), Community and response plans and DRR program implementation. In the Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

study area, the UPs formed disaster management committees, DISCUSSION AND CONCLSUION called Union-Parishad Disaster Management Committees The fundamental linkage between social learning and community (UMDC) to carry out these activities. None of the UMDC resilience to NTEE and associated disasters was our point of entry members were elders, nor have they been consulted in the formal into this investigation. We took a critical stance on the processes. The UMDCs consisted mostly of local elites and programmatic intervention of community resilience and social representatives from different professional groups who learning and argued that local knowledge, beliefs, practices, and collaborated with NGOs to implement planning and training SM are crucial elements in the social learning process for building activities. An examination of training manuals and planning community resilience to NTEEs and disasters. Our findings are documents revealed the total absence of ILK and SM in them. A novel in three respects: (i) alternative possibilities for and local representative from an NGO explained: notions of resilience are undermined when ILK is unevenly folded into programmatic interventions for social learning and resilience While working with UP and UDMC, we mostly work building, which can paradoxically make communities more with the guidelines provided by the national and district vulnerable to environmental extremes; (ii) by combining various level administration and, when needed, consult with the dimensions of ILK with SM and official warnings and local elites as representatives from the community. information, local knowledge holders can successfully generate However, as community members trust us, in some accurate EWs to reduce risks due to cyclones; and (iii) the exceptional cases, we consult with community members asymmetrical distribution of ILK in terms of occupation, gender, at the grass-root level. and age across communities highlights the value of social learning During fieldwork we observed that, through a community-based for system-wide (i.e., community) learning and building participatory approach, NGOs sometimes attempt to incorporate community resilience to NTEEs. local needs and knowledge in the disaster planning and response These findings have serious implications for the development and process. For example, prior to preparing local risk and resources implementation of future strategies for reducing the increasing mapping, an NGO (NSS) organized FGD meetings with risks posed by climate-change-induced hydrometeorological community people. In the deliberations, people identified the extreme events like tropical cyclones (Woodruff et al. 2013, usual timing of NTEE, including saline water intrusion during Marsooli et al. 2019, Uriarte et al. 2019). To cite some recent the winter season (December–February) and cyclones and storm examples, Hurricane Harvey in the southern United States of surges during the monsoon (April and November). Indigenous America brought with it unprecedented volumes of rainwater for and local knowledge thus received some attention and was being which most communities were not prepared, and the later incorporated into such planning. In addition, in the participatory Category 5 Hurricane Irma lasted far longer than any storm of learning platforms, NGO workers collected information and its size in history (Rahmstorf 2017). We found that community knowledge of local people’s (traumatic) memory of cyclones and people have rich stocks of ILK that help in generating EWs to storm surges to lay the groundwork for disseminating scientific– reduce disaster risks and build resilience to coastal cyclones. technocratic knowledge on EW and textbook ideas of However, climate-induced shocks could still appear as a surprise preparedness. for community people. Cyclone Sidr struck as a surprise to many Our research findings reveal that community people can generate coastal communities as they were unable to predict the storm EWs from various dimensions of ILK to reduce risk of loss and surge danger with the celestial ILK dimension. Some studies, damage and enhance resilience to coastal cyclones. However, in however, suggest that some communities are capable of adapting formal social learning processes, such forms of knowledge are well to climate extremes that surpass the parameters predicted by generally subjugated to a scientific–technocratic form of IPCC scenario-building models (Nyong et al. 2007), whereas knowledge on EW, preparedness, and evacuation. Therefore, other communities are struggling to adapt with their ILK (Lebel community members were being taught only the meanings of 2013, Kagunyu et al. 2016). In this regard, several authors rightly official EW signs and signals. In the absence of prior positive SM argue for collaborative knowledge production for building and with prevailing traumatic SM of Sidr, scientific–technocratic resilience and facilitating effective disaster management forms of knowledge appeared attractive to young and middle- (Srivastava 2012, Sitas et al. 2016, Rodela and Swartling 2019). aged participants. By integrating ILK with scientific knowledge to formulate and disseminate EWs, the risks posed by climate-induced disaster The key to translating knowledge into action for building shocks can be substantially reduced (Fig. 6). resilience is the presence of positive SM, which is mostly held by elders. Moreover, elders successfully combined official warnings The problem of ILK and technical–scientific knowledge with ILK to generate their own EWs, and those who listened to integration raises some critical questions: how can this potential them were able to avoid loss and damage. Because the social integration take place, and what are the power–knowledge learning platforms systematically exclude elders, important dynamics within an institutional context that in turn shape social dimensions of ILK, positive SM to translate this knowledge into learning processes and resilience outcomes? Most proposals and action, and elders’ capacity for integrating scientific–technocratic efforts have hitherto sought to integrate ILK with scientific knowledge with ILK are frequently excluded as well. People knowledge via community-based participatory approaches (Tran therefore are likely to ignore EWs from official sources (as and Rodela 2019). Nadasdy (1999, 2005), however, cautions discussed earlier) and from ILK, and suffer losses and damages regarding an integration process where ILK is unevenly folded as a result. Thus, the imposition of scientific–technocratic into formal institutional practices, and as a consequence, efforts knowledge and subjugation of ILK by community-based to build and/or enhance resilience through community-based resilience programming may in many cases enhance vulnerability interventions may have the unintended consequence of eroding rather than resilience. rather than enhancing resilience. Our study demonstrates that Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

NGOs tend to consider ILK in community-based resilience highlights the empirical evidence that formal institutions do not programming in some cases merely to lay the groundwork for recognize the significance of ILK, and therefore operate on the disseminating scientific–technocratic knowledge on EW and basis of predesigned resilience frameworks defined by nonlocals preparedness. and external policy makers (Fig. 6). Indigenous and local knowledge is thus excluded from programmatic social learning Fig. 6. Interconnections between communities’ inner strengths processes, leading to the subjugation of local community and formal institutional measures, and their effects upon DRR members to externally defined conceptions of resilience and and community resilience. consequently reducing learning from such events to biopolitical rationalities (Hofmann 2014). Furthermore, because these social learning platforms systematically exclude elders and other local sources of ILK, key dimensions of ILK and the positive SM necessary for translating knowledge into action are excluded too. Due to this lack of cogenerated knowledge and reliable information, people are likely to ignore EWs from official sources and from ILK, and consequently suffer loss and damage. The imposition of scientific–technocratic knowledge and subjugation of ILK by neoliberal community-based resilience programming may enhance vulnerability rather than building communities’ inner strengths and capabilities. The relationship between ILK and SM is often not clearly delineated. Most studies on DRR and resilience either address ILK or SM separately or only mention one in relation to the other (cf. Garde-Hansen et al. 2017, Moreno et al. 2019, Setten and Lein 2019). Our detailed examination of ILK and SM emphasizes that SM is a necessary condition for translating knowledge into action (Fig. 6). We also observed that community members’ responses to disaster risks are shaped by the nature of their SM, despite having a rich stock of ILK. Such gaps are also known as the “temporal variability in hazardscapes” (de Vries 2011). Madsen and Mullan (2013) argue that as NTEEs and disasters are episodic events, each event is remembered and considered as an isolated event rather than as part of a larger trend. In our study communities, people were familiar with gusty winds and monsoon rain but they had no prior SM of severe cyclones and associated It can, therefore, be argued that the goal of social learning storm-surge-induced flooding, and therefore, the impacts of processes should be to nurture and build upon local strengths for Cyclone Sidr were a complete surprise for most communities. This developing communities’ own conceptualization and components resonates with Berkes and Ross’s (2013) observation that the of resilience. In this regard, Berkes and Folke (1998) argue that a attributes of community resilience tend to differ depending on resilient system has inbuilt “social mechanisms” based on local the types of shocks experienced (e.g., floods or wildfire). However, knowledge that act as buffers against disturbances and maintain the presence of positive SM (mostly held by elders) made a community resilience. Rather than solely relying on conventional significant difference in terms of risk reduction and resilience, interventionist strategies, the goal should be to recognize the which has also been documented in other contexts (e.g., Berkes power of human agency, dignity, and capability at the local level, 2007, Osterhoudt 2018) (Fig. 6). and to provide and cogenerate the resources necessary so that communities can make their own choices for building and In this article, we attempted to integrate scholarship on social enhancing resilience (Kevin and Jonathan 2015, Evans and Reid learning, community resilience, and ILK, and address the issue 2014). Our study in coastal Bangladesh provides evidence that the of gaps between the community resilience literature and some of local community have adopted modern technologies and its social science critics (cf. Grove 2014b, 2014a, Hill and Larner knowledge in their own way to generate EWs to coastal cyclones. 2017). Regarding the former, we found that using ILK as the point Hilhorst et al. (2015) have similarly found that indigenous people of convergence between the social learning and the community proactively adopt modern technologies to adapt with changing resilience literature helps to substantially improve our circumstances. understanding of actual social learning and resilience building processes. Regarding the engagement of critical social theories Although social learning is a process of collaborative knowledge with the current community resilience literature, we hold that the production, and the knowledge produced through these processes former present a similar argument, i.e., in building resilience, the shapes resilience pathways (Barrios 2016, Boyd et al. 2014), state focus should be on building upon a community’s strengths rather agencies and formal institutions (e.g., NGOs) are increasingly than simply correcting its perceived weaknesses. Further efforts taking unilateral, interventionist approaches to enhancing are required for bridging these literatures, including critical resilience to natural hazards (Grove 2013a). Our critical stance engagement and interaction, dialog and deliberation, and on the role of formal institutions in the social learning process integration and knowledge cogeneration, as one anonymous Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

reviewer succinctly states that, at present the “two strands of LITERATURE CITED literature ... often talk past each other,” and this needs to be Acharya, A., and A. Prakash. 2019. When the river talks to its transformed into “meaningful interaction.” people: local knowledge-based flood forecasting in Gandak River basin, India. Environmental Development 31:55–67. https://doi. In this article, we limited our investigation to whether the org/10.1016/j.envdev.2018.12.003 established, formal institutions consider local voices, learning, ILK, and SM in social learning processes, and whether these Adger, W. N., T. A. Benjaminsen, K. Brown, and H. Svarstad. processes are characterized mostly by a top-down structure. There 2001. Advancing a political ecology of global environmental is also a need to investigate the synergies among learning at discourses. Development and Change 32(4):681–715. https://doi. different institutional levels and mechanisms for scaling-up org/10.1111/1467-7660.00222 learning from community and lower level institutions to higher Adger, W. N., T. P. Hughes, C. Folke, S. R. Carpenter, and J. level institutions. Although we specifically focused on the role of Rockstrom. 2005. Social–ecological resilience to coastal disasters. social learning in shaping community resilience and the role of Science 309(1036):1036–1039. https://doi.org/10.1126/science.1112122 locals as active agents in such learning processes, further research needs to be carried out on how other forms of learning (e.g., Agrawal, A. 2005. Environmentality: community, intimate transformative learning) function to shape community resilience. government, and the making of environmental subjects in Further research on how worldviews, beliefs, values, and other Kumaon, India. Current Anthropology 46(2):161–190. https://doi. cultural factors shape learning and resilience building processes org/10.1086/427122 is also needed. Specifically, the views, meanings, and interpretations of community people concerning resilience to Armitage, D., M. Marschke, and R. Plummer. 2008. Adaptive co- NTEEs and disasters and the roles they can play in the management and the paradox of learning. Global Environmental coproduction of pertinent knowledge need further attention. As Change 18(1):86–98. https://doi.org/10.1016/j.gloenvcha.2007.07.002 an anonymous reviewer appropriately suggested, “it is important to consider potential tensions or clashes in attitudes and Baird, J., R. Plummer, C. Haug, and D. Huitema. 2014. Learning knowledge among [heterogeneous] indigenous people and effects of interactive decision-making processes for climate inequities within local power structures” in the analysis of efforts change adaptation. Global Environmental Change 27(1):51–63. to integrate ILK and scientific knowledge. We agree that such https://doi.org/10.1016/j.gloenvcha.2014.04.019 diversities and complexities should not be overlooked, but rather Barrios, R. E. 2016. Resilience: a commentary from the vantage demand indepth examination. point of anthropology. Annals of Anthropological Practice 40 (1):28–38. https://doi.org/10.1111/napa.12085 Responses to this article can be read online at: Beilin, R., and C. Wilkinson. 2015. Introduction: governing for https://www.ecologyandsociety.org/issues/responses. urban resilience. Urban Studies 52(7):1205–1217. https://doi. php/12107 org/10.1177/0042098015574955 Berkes, F. 2007. Understanding uncertainty and reducing vulnerability: lessons from resilience thinking. Natural Hazards Acknowledgments: 41(2):283–295. https://doi.org/10.1007/s11069-006-9036-7 This work was supported by a Social Science and Humanities Berkes, F. 2018. Context of traditional ecological knowledge. Research Council (SSHRC) Insight Grant [Grant # 435-2018-552]. Pages 1–20 in F. Berkes. Sacred ecology. Fourth edition. The authors are thankful to SSHRC for its financial support. The Routledge, New York, New York, USA. https://doi. first author is indebted to Canada Research Chair [CRC Tier I] org/10.4324/9781315114644-1 Fund on Community-Based Resources Management and the Berkes, F., and C. Folke. 1998. Linking social and ecological University of Manitoba for Graduate Fellowship award for systems for resilience and sustainability. Pages 1–25 in F. Berkes supporting his Ph.D. studies. The authors are thankful to Fikret and C. Folke, editors. Linking social and ecological systems: Berkes (Distinguished Professor Emeritus, Natural Resources management practices and social mechanisms for building Institute, University of Manitoba, Canada) for his insightful resilience. Cambridge University Press, Cambridge, UK. comments on the manuscript, and to Sultan M Shuvon (Shahjalal University of Science and Technology, Bangladesh) for his help Berkes, F., C. Folke. 2002. Back to the future: ecosystem dunamics preparing the study map. The authors are also grateful to the and local knowledge. Pages 121–146 in L. H. Gunderson and C. community people for their spontaneous participation in the S. Holling, editors. Panarchy: understanding transformations in research process. human and natural systems. Island Press, Washington, D.C., USA.

Berkes, F., and H. Ross. 2013. Community resilience: toward an Data Availability: integrated approach. Society and Natural Resources 26(1):5–20. The data/code that support the findings of this study are available https://doi.org/10.1080/08941920.2012.736605 on request from the corresponding author, [CEH]. The data/code Bockstael, E. 2017. Critical capacity development: an action are not publicly available due to institutional agreement-based research approach in coastal Brazil. World Development 94:336– restrictions, e.g., their containing information that could 345. https://doi.org/10.1016/j.worlddev.2017.01.017 compromise the privacy of research participants. Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Boillat, S., and F. Berkes. 2013. Perception and interpretation of Christensen, L., and N. Krogman. 2012. Social thresholds and climate change among Quechua farmers of Bolivia. Ecology and their translation into social–ecological management practices. Society 18(4): 21. https://dx.doi.org/10.5751/ES-05894-180421 Ecology and Society 17(1): 5. http://dx.doi.org/10.5751/ ES-04499-170105 Bonilla, Y. 2020. The coloniality of disaster: race, empire, and the temporal logics of emergency in Puerto Rico, USA. Political Cleaver, F. 2007. Understanding agency in collective action. Geography 78:102181. https://doi.org/10.1016/j.polgeo.2020.102181 Journal of Human Development 8(2):223–244. http://dx.doi. org/10.1080/14649880701371067 Boyd, E., J. Ensor, V. C. Broto, and S. Juhola. 2014. Collier, S. J. 2009. Topologies of power: Foucault’s analysis of Environmentalities of urban climate governance in Maputo, political government beyond ‘governmentality.’ Theory, Culture Mozambique. Global Environmental Change 26(1):140–151. and Society 26(6):78–108. doi: http://dx.doi.org/10.1177/0263276409347694 https://doi.org/10.1016/j.gloenvcha.2014.03.012 Brammer, H. 1990. Floods in Bangladesh: geographical Cote, M., and A. J. Nightingale. 2011. Resilience thinking meets background to the 1987 and 1988 floods. The Geographical social theory: situating social change in socio-ecological systems Journal 156(1):12–22. https://doi.org/https://doi.org/10.2307/635431 (SES) research. Progress in Human Geography 36(4):475–489. http://dx.doi.org/10.1177/0309132511425708 Briggs, J., and J. Sharp. 2004. Indigenous knowledges and Creswell, J. W. 2007. Qualitative inquiry and research design: development: a postcolonial caution. Third World Quarterly 25 choosing among five approaches. First edition. Sage Publications, (4):661–676. https://doi.org/10.1080/01436590410001678915 Los Angeles, California, USA. Brown, K. 2016. Resilience, development and global change. Creswell, J. W. 2013. Qualitative inquiry and research design: Routledge, London, UK and New York, New York, USA. https:// choosing among five approaches. Third edition. Sage Publications, doi.org/10.4324/9780203498095 Thousand Oaks, California, USA. Brown, K., and E. Westaway. 2011. Agency, capacity, and Creswell, J. W. 2014. Research design: qualitative, quantitative, and resilience to environmental change: lessons from human mixed methods approaches. Fourth edition. Sage Publications, development, well-being, and disasters. Annual Review of Thousand Oaks, California, USA. Environment and Resources 36(1):321–342. https://doi.org/10.1146/ annurev-environ-052610-092905 Cretney, R. 2014. Resilience for whom? Emerging critical geographies of socio-ecological resilience. Geography Compass 8 Bulley, D. 2013. Producing and governing community (through) (9):627–640. https://doi.org/10.1111/gec3.12154 resilience. Politics 33(4):265–275. https://doi.org/10.1111/1467-9256.12025 Cutter, S. L., L. Barnes, M. Berry, C. Burton, E. Evans, E. Tate, Cannon, T., and D. Müller-Mahn. 2010. Vulnerability, resilience and J. Webb. 2008. A place-based model for understanding and development discourses in context of climate change. Natural community resilience to natural disasters. Global Environmental Hazards 55(3):621–635. https://doi.org/10.1007/s11069-010-9499-4 Change 18(4):598–606. https://doi.org/10.1016/j.gloenvcha.2008.07.013

Carr, E. R. 2019. Properties and projects: reconciling resilience de Vries, D.H., 2011. Temporal vulnerability in hazardscapes: and transformation for adaptation and development. World flood memory-networks and referentiality along the North Development 122:70–84. https://doi.org/10.1016/j.worlddev.2019.05.011 Carolina Neuse River (USA). Global Environmental Change 21:154–164. https://doi.org/10.1016/j.gloenvcha.2010.09.006 Chandler, D. 2018. Ontopolitics in the anthropocene:an Dekens, J. 2007. Local knowledge for disaster preparedness: a introduction to mapping, sensing and hacking, Routledge, London, literature review. International Centre for Integrated Mountain UK and New York, New York, USA. https://doi.org/https://doi. Development (ICIMOD), Hill Side Press (P) Ltd., Kathmandu, org/10.4324/9780203703434 Nepal. Chandler, D. 2019. Resilience and the end(s) of the politics of Dunstan, P. K., S. D. Foster, E. King, J. Risbey, T. J. O’Kane, D. adaptation. Resilience: International Policies, Practices and Monselesan, A. J. Hobday, J. R. Hartog, and P. A. Thompson. Discourses 7(3):304–313. https://doi.org/10.1080/21693293.2019.1605660 2018. Global patterns of change and variation in sea surface temperature and chlorophyll a. Scientific Reports 8(1):1–9. https:// Choudhury, M. U. I., and C. E. Haque. 2016. “We are more scared doi.org/10.1038/s41598-018-33057-y of the power elites than the floods”: adaptive capacity and resilience of wetland community to flash flood disasters in Ensor, J., and B. Harvey. 2015. Social learning and climate change Bangladesh. International Journal of Disaster Risk Reduction adaptation: evidence for international development practice. 19:145–158. https://doi.org/10.1016/j.ijdrr.2016.08.004 Wiley Interdisciplinary Reviews: Climate Change 6(5):509–522. https://doi.org/10.1002/wcc.348 Choudhury, M.-U.-I., M. S. Uddin, and C. E. Haque. 2019. “Nature brings us extreme events, some people cause us prolonged Evans, B., and J. Reid. 2013. Dangerously exposed: the life and sufferings”: the role of good governance in building community death of the resilient subject. Resilience: International Policies, resilience to natural disasters in Bangladesh. Journal of Practices and Discourses 1(2):83–98. https://doi.org/10.1080/216­ Environmental Planning and Management 62(10):1761–1781. 93293.2013.770703 https://doi.org/10.1080/09640568.2018.1513833 Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Evans, B., and J. Reid. 2014. Resilient life: the art of living European Commission. [online] URL: https://reliefweb.int/sites/ dangerously. Wiley, New York, New York, USA. reliefweb.int/files/resources/F2FDFF067EF49C8DC12574DC00455142- Full_Report.pdf Fabinyi, M., L. E. Foale, S. J., L. Evans, S. J. S. J. Foale, C. Reflections, and C. Debate. 2014. Social–ecological systems, Grove, K. 2013a. Hidden transcripts of resilience: power and social diversity , and power: insights from anthropology and politics in Jamaican disaster management. Resilience: political ecology. Ecology and Society 19(4): 28. http://dx.doi. International Policies, Practices and Discourses 1(3):193–209. org/10.5751/ES-07029-190428 https://doi.org/10.1080/21693293.2013.825463 Faulkner, L., K. Brown, and T. Quinn. 2018. Analyzing Grove, K. 2013b. On resilience politics: from transformation to community resilience as an emergent property of dynamic social– subversion. Resilience: International Policies, Practices and ecological systems. Ecology and Society 23(1): 24. https://doi. Discourses 1(2):146–153. https://doi.org/10.1080/21693293.2013.804661 org/10.5751/ES-09784-230124 Grove, K. 2014a. Agency, affect, and the immunological politics Folke, C., S. R. Carpenter, B. Walker, M. Scheffer, T. Chapin, and of disaster resilience. Environment and Planning D: Society and J. Rockström. 2010. Resilience thinking: integrating resilience, Space 32(2):240–256. https://doi.org/10.1068/d4813 adaptability and transformability. Ecology and Society 15(4): 20. https://doi.org/http://www.ecologyandsociety.org/vol15/iss4/art20/ Grove, K. 2014b. Biopolitics and adaptation: governing socio- ecological contingency through climate change and disaster Folke, C., J. Colding, and F. Berkes. 2003. Synthesis: building studies. Geography Compass 8(3):198–210. https://doi.org/10.1111/ resilience and adaptive capacity in social–ecological systems. gec3.12118 Pages 352–387 in F. Berkes, J. Colding, and C. Folke, editors. Grove, K. 2018. Resilience as subjugated knowledge. Pages 65– Navigating social–ecological systems: building resilience for 88 in K. Grove. Resilience. Routledge, London, UK and New complexity and change. Cambridge University Press, Cambridge, York, New York, USA. https://doi.org/10.4324/9781315661407-3 UK. https://doi.org/10.1017/CBO9780511541957.020 Folke, C., T. Hahn, P. Olsson, and J. Norberg. 2005. Adaptive Gunderson, L. 2010. Ecological and human community resilience governance of social-ecological systems. Annual Review of to natural disasters. Ecology and Society 15(2): 18. https://doi. Environment and Resources 30(1):441–473. https://doi.org/10.1146/ org/10.5751/ES-03381-150218 annurev.energy.30.050504.144511 Haque, C. E., M.-U.-I. Choudhury, and M. S. Sikder. 2018. Foucault, M. 1978. The history of sexuality. Volume 1. Pages 40– “Events and failures are our only means for making policy 42 in M. Foucault and S. Mills. 2005, Routledge critical thinkers changes”: learning in disaster and emergency management essential guides for literary studies. Routledge, London, UK. policies in Manitoba, Canada. Natural Hazards 98:137–162 . https://doi.org/10.1007/s11069-018-3485-7 Foucault, M. 1980. Power/knowledge. Pages 67–80 in M. Foucault and S. Mills. 2005, Routledge critical thinkers essential guides for Haque, C. E., and M. S. Uddin. 2013. Disaster management literary studies. Routledge, London, UK. https://doi. discourse in Bangladesh: a shift from post-event response to the org/10.4324/9781003060963-10 preparedness and mitigation approach through institutional partnerships. Pages 33–44 in J. Tiefenbacher. editor. Approaches Gaillard, J. C. 2010. Vulnerability, capacity and resilience: to disaster management: examining the implications of hazards, Perspectives for climate and development policy. Journal of emergencies and disasters. IntechOpen, London, UK. https://doi. International Development 22:218–232. https://doi.org/10.1002/ org/https://doi.org/10.5772/54973 jid.1675 Henderson-Sellers, A., H. Zhang, G. Berz, K. Emanuel, W. Gray, Garde-Hansen, J., L. McEwen, A. Holmes, and O. Jones. 2017. C. Landsea, G. Holland, J. Lighthill, S. L. Shieh, P. Webster, and Sustainable flood memory: remembering as resilience. Memory K. McGuffie. 1998. Tropical cyclones and global climate change: Studies 10(4):384–405. https://doi.org/10.1177/1750698016667453 a post-IPCC assessment. Bulletin of the American Meteorological Society 79(1):19–38. https://doi.org/10.1175/1520-0477(1998) Gómez-Baggethun, E., E. Corbera, and V. Reyes-García. 2013. 079<0019:TCAGCC>2.0.CO;2 Traditional ecological knowledge and global environmental change: research findings and policy implications. Ecology and Hernandez, O., J. Jouanno, V. Echevin, and O. Aumont. 2017. Society 18(4): 72. http://dx.doi.org/10.5751/ES-06288-180472 Modification of sea surface temperature by chlorophyll concentration in the Atlantic upwelling systems. Journal of Gómez-Baggethun, E., V. Reyes-García, P. Olsson, and C. Geophysical Research: Oceans 122:2647–2651. https://doi.org/ Montes. 2012. Traditional ecological knowledge and community https://doi.org/10.1002/2016jc012330 resilience to environmental extremes: a case study in Doñana, SW Spain. Global Environmental Change 22(3):640–650. http://dx. Hewer, C. J., and M. Kut. 2010. Historical legacy, social memory doi.org/10.1016/j.gloenvcha.2012.02.005 and representations of the past within a Polish community. Memory Studies 3(1):18–32. https://doi.org/10.1177/1750698009348292 Government of Bangladesh (GoB). 2008. Cyclone Sidr in Bangladesh: damage, loss, and needs assessment for disaster Hilhorst, D., J. Baart, G. van der Haar, and F. M. Leeftink. 2015. recovery and reconstruction. A report prepared by the Is disaster “normal” for indigenous people? Indigenous Government of Bangladesh assisted by the international knowledge and coping practices. Disaster Prevention and development community with financial support from the Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Management 24(4):468–483. https://doi.org/https://doi.org/10.1108/ Komac, B. 2009. Social memory and geographical memory of dpm-02-2015-0027 natural disasters. Acta geographica Slovenica 49(1):199–226. https://doi.org/10.3986/ags49107 Hill, L. J., and W. Larner. 2017. The resilient subject. Pages 263– 281 in V. Higgins and W. Larner, editors. Assembling neoliberalism: Kossin, J. P. 2017. Hurricane intensification along United States expertise, practices, subjects. Springer Nature, Berlin/Heidelberg, coast suppressed during active hurricane periods. Nature 541 Germany. https://doi.org/10.1057/978-1-137-58204-1 (7637):390–393. https://doi.org/10.1038/nature20783 Hofmann, M. 2014. Reduced learning processes due to Lebel, L. 2013. Local knowledge and adaptation to climate biopolitical patterns of interpretation: Michel Foucault and the change in natural resource-based societies of the Asia-Pacific. contamination disaster. Pages 253–271 in H. Egner, M. Schorch, Mitigation and Adaptation Strategies for Global Change 18 and M. Voss, editors. Learning and calamities: practices, (7):1057–1076. https://doi.org/10.1007/s11027-012-9407-1 interpretations, patterns. Routledge, London, UK and New York, MacKinnon, D., and K. D. Derickson. 2013. From resilience to New York, USA. https://doi.org/10.4324/9780203794678 resourcefulness: a critique of resilience policy and activism. Holling, C. S. 2004. From complex region to complex worlds. Progress in Human Geography 37(2):253–270. https://doi. Ecology and Society 9(1): 11. http://www.ecologyandsociety.org/ org/10.1177/0309132512454775 vol9/iss1/art11 https://doi.org/10.5751/ES-00612-090111 Madsen, W., and C. O. Mullan. 2013. Responding to disaster Hooli, L. J. 2016. Resilience of the poorest: coping strategies and applying the lens of social memory. Australian Journal of indigenous knowledge of living with the floods in northern Communication 40(1):57–70. https://search.informit.com.au/ Namibia. Regional Environmental Change 16(3):695–707. https:// documentSummary;dn=430871383389493;res=IELAPA doi.org/10.1007/s10113-015-0782-5 Magis, K. 2010. Community resilience: an indicator of social International Panel on Climate Change (IPCC). 2014. Summary sustainability. Society and Natural Resources 23(5):401–416. for policymakers. Climate Change 2014: Impacts, Adaptation and http://dx.doi.org/10.1080/08941920903305674 Vulnerability - Contributions of the Working Group II to the Fifth Maneesha, K., V. S. N. Murty, M. Ravichandran, T. Lee, W. Yu, Assessment Report. IPCC, Geneva, Switzerland. [online] URL: and M. J. McPhaden. 2012. Upper ocean variability in the Bay of https://www.ipcc.ch/site/assets/uploads/2018/11/SR1.5_SPM_Low_Res. Bengal during the tropical cyclones Nargis and Laila. Progress in pdf Oceanography 106:49–61. https://doi.org/10.1016/j.pocean.2012.06.006 Jacobson, M., and C. Rugeley. 2007. Community-based participatory research: group work for social justice and Marchezini, V. 2015. The biopolitics of disaster: power, community change. Social Work with Groups 30(4):21–39. https:// discourses, and practices. Human Organization 74(4):362–371. doi.org/10.1300/J009v30n04_03 https://doi.org/10.17730/0018-7259-74.4.362 Johannessen, Å., and T. Hahn. 2013. Social learning towards a Marsooli, R., N. Lin, K. Emanuel, and K. Feng. 2019. Climate more adaptive paradigm? Reducing flood risk in Kristianstad change exacerbates hurricane flood hazards along U.S. Atlantic municipality, Sweden. Global Environmental Change 23(1):372– coast in spatially varying patterns. Nature Communications 10 381. https://doi.org/10.1016/j.gloenvcha.2012.07.009 (1):1–9. https://doi.org/10.1038/s41467-019-11755-z Jordan, J. C. 2018. Deconstructing resilience: why gender and Melo Zurita, M. de L., B. Cook, L. Harms, and A. March. 2015. power matter in responding to climate stress in Bangladesh. Towards new disaster governance: subsidiarity as a critical tool. Climate and Development 11(2):167–179 https://doi. Environmental Policy and Governance 25(6):386–398. https://doi. org/10.1080/17565529.2018.1442790 org/10.1002/eet.1681 Kagunyu, A., S. Wandibba, and J. G. Wanjohi. 2016. The use of Melo Zurita, M. de L., B. Cook, D. C. Thomsen, P. G. Munro, T. indigenous climate forecasting methods by the pastoralists of F. Smith, and J. Gallina. 2018. Living with disasters: social capital northern Kenya. Pastoralism 6(7):1–6. https://doi.org/10.1186/ for disaster governance. Disasters 42(3):571–589. https://doi. s13570-016-0054-0 org/10.1111/disa.12257 Kelman, I., J. Mercer, and J. Gaillard. 2012. Indigenous Mertens, D. M. 2007. Transformative paradigm: mixed methods knowledge and disaster risk reduction. Geography 97(1):12–21. and social justice. Journal of Mixed Methods Research 1(3):212– https://doi.org/10.1080/00167487.2012.12094332 225. https://doi.org/10.1177/1558689807302811 Kevin, G., and P. Jonathan. 2015. Assemblage thinking and Mertens, D.M., 2009. Transformative research and evaluation. participatory development: potentiality, ethics, biopolitics. Guilford Press, New York, New York, USA. Geography Compass 9(1):1–13. https://doi.org/10.1111/gec3.12191 Mertens, D. M. 2010. Transformative mixed methods research. Qualitative Inquiry 16(6):469–474. https://doi.org/10.1177/1077­ Khan, T. M. A., O. P. Singh, and M. S. Rahman. 2000. Recent 800410364612 sea level and sea surface temperature trends along the bangladesh coast in relation to the frequency of intense cyclones. Marine Mertens, D. M. 2012. Transformative mixed methods: addressing Geodesy 23(2):103–116. https://doi.org/10.1080/01490410050030670 inequities. American Behavioral Scientist 56(6):802–813. https:// doi.org/10.1177/0002764211433797 Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Ministry of Disaster Management and Relief (MoDMR). 2017. and cultural responses. The Rowman and Littlefield Publishing National plan for disaster management (2016–2020): Building Groups, Inc., Lanham, Maryland, USA. resilience for sustainable human development. MoDMR, Dhaka, Pugh, J. 2013. Speaking without voice: participatory planning, Bangladesh. acknowledgment, and latent subjectivity in Barbados. Annals of Moreno, J., A. Lara, and M. Torres. 2019. Community resilience the Association of American Geographers 103(5):1266–1281. in response to the 2010 tsunami in Chile: the survival of a small- https://doi.org/10.1080/00045608.2012.706571 scale fishing community. International Journal of Disaster Risk Rahmstorf, S. 2017. Rising hazard of storm-surge flooding. Reduction 33(November 2018):376–384. https://doi.org/10.1016/ Proceedings of the National Academy of Sciences 114 j.ijdrr.2018.10.024 (45):201715895. https://doi.org/10.1073/pnas.1715895114 Nadasdy, P. 1999. The politics of TEK: power and the Reed, M., A. A. Evely, G. Cundill, I. Fazey, J. Glass, A. Laing, J. “integration” of knowledge. Arctic Anthropology 36(1–2):1–18. Newig, B. Parrish, C. Prell, C. Raymond, L. Stringer, and M. S. https://www.jstor.org/stable/40316502 Reed Evely, A. C., Cundill, G., Fazey, I., Glass, J., Laing, A., Nadasdy, P. 2005. The anti-politics of TEK: the Newig, J., Parrish, B., Prell, C., Raymond, C. and Stringer, L. C. institutionalization of co-management discourse and practice. 2010. What is social learning? Ecology and Society 15(4): r1. Anthropologica 47(2):215–232. https://www.jstor.org/stable/25606237 https://doi.org/10.5751/es-03564-1504r01 Rodela, R. 2011. Social learning and natural resource Nyong, A., F. Adesina, and B. Osman Elasha. 2007. The value of management : the emergence of three research perspectives. indigenous knowledge in climate change mitigation and Ecology and Society 16(4): 30. https://doi.org/10.5751/ adaptation strategies in the African Sahel. Mitigation and ES-04554-160430 Adaptation Strategies for Global Change 12(5):787–797. https:// www.doi.org/10.1007/s11027-007-9099-0 Rodela, R. 2013. The social learning discourse: trends, themes and interdisciplinary influences in current research. Environmental Olick, J. K. 2016. The poverty of resilience: on memory, meaning, Science and Policy 25:157–166. https://doi.org/10.1016/j. and well-being. Memory Studies 9(3):315–324. https://doi. envsci.2012.09.002 org/10.1177/1750698016645265 Rodela, R., and Å. G. Swartling. 2019. Environmental governance Oliver-Smith, A. 2002. Theorizing disasters: nature, power, and in an increasingly complex world: reflections on transdisciplinary culture. Pages 23–48 in S. M. Hoffman and A. Oliver-Smith, collaborations for knowledge coproduction and learning. editors. Catastrophe and culture: the anthropology of disaster. Environmental Policy and Governance 29(2):83–86. https://doi. School of American Research Press, Santa Fe, New Mexico, USA. org/10.1002/eet.1842 https://doi.org/https://doi.org/10.4000/lhomme.2445 Roy, T. 2012. Natural disaster and Indian history. Oxford Osterhoudt, S. 2018. Remembered resilience : oral history University Press, New Delhi, India. narratives and community resilience in agroforestry systems. Renewable Agriculture and Food Systems 33(3):252–255. https:// Rumbach, A., and D. Foley. 2014. Indigenous institutions and doi.org/https://doi.org/10.1017/s1742170517000679 their role in disaster risk reduction and resilience: evidence from the 2009 tsunami in American Samoa. Ecology and Society 19(1): Ostrom, E. 2008. Institutions and the environment. Economic 19 https://doi.org/10.5751/ES-06189-190119 Affairs 28(3):24–31. https://doi.org/10.1111/j.1468-0270.2008.00840. x Secor, D. H., F. Zhang, M. H. P. O’Brien, and M. Li. 2019. Ocean destratification and fish evacuation caused by a Mid-Atlantic Pahl-Wostl, C. 2006. The importance of social learning in tropical storm. ICES Journal of Marine Science 76(2):573–584. restoring the multifunctionality of rivers and floodplains. Ecology https://doi.org/10.1093/icesjms/fsx241 and Society 11(1): 10 [online] URL: http://www.ecologyandsociety. org/vol11/iss1/art10/ https://doi.org/10.5751/ES-01542-110110 Setten, G., and H. Lein. 2019. “We draw on what we know anyway”: the meaning and role of local knowledge in natural Pahl-Wostl, C. 2009. A conceptual framework for analysing hazard management. International Journal of Disaster Risk adaptive capacity and multi-level learning processes in resource Reduction 38(May):101184. https://doi.org/10.1016/j.ijdrr.2019.101184 governance regimes. Global Environmental Change 19(3):354–365. https://doi.org/10.1016/j.gloenvcha.2009.06.001 Sitas, N., B. Reyers, G. Cundill, H. E. Prozesky, J. L. Nel, and K. Paul, B. K. 2009. Why relatively fewer people died? The case of J. Esler. 2016. Fostering collaboration for knowledge and action Bangladesh’s cyclone sidr. Natural Hazards 50(2):289–304. in disaster management in South Africa. Current Opinion in https://doi.org/10.1007/s11069-008-9340-5 Environmental Sustainability 19(December 2015):94–102. https:// doi.org/10.1016/j.cosust.2015.12.007 Pelling, M. 2011. Adaptation to climate change: from resilience to tranformation. Routledge, London, UK and New York, New Spampinato, C., S. Palazzo, B. Boom, J. Van Ossenbruggen, I. York, USA. https://doi.org/10.4324/9780203889046 Kavasidis, R. Di Salvo, F. P. Lin, D. Giordano, L. Hardman, and R. B. Fisher. 2014. Understanding fish behavior during typhoon Pfister, C. 2009. Learning from nature-induced disasters: events in real-life underwater environments. Multimedia Tools and theoretical considerations and case studies from Western Europe. Applications 70(1):199–236. https://doi.org/10.1007/s11042-012-1101-5 Pages 17–40 in C. Mauch and C. Pfister, editors. Natural disasters Ecology and Society 26(1): 5 https://www.ecologyandsociety.org/vol26/iss1/art5/

Srivastava, S. K. 2012. Managing indigenous and scientific knowledge for resilience building. Journal of Advances in Management Research 9(1):45–63. https://doi.org/10.1108/0972­ 7981211225644 Thiessen, C., and S. Byrne. 2017. Proceed with caution: research production and uptake in conflict-affected countries. Journal of Peacebuilding and Development 13:1–15. https://doi. org/10.1080/15423166.2017.1401486 Thomalla, F., and R. K. Larsen. 2010. Resilience in the context of tsunami early warning systems and community disaster preparedness in the indian ocean region. Environmental Hazards 9(3):249–265. https://doi.org/10.3763/ehaz.2010.0051 Tidball, K. G., M. E. Krasny, E. Svendsen, L. Campbell, and K. Helphand. 2010. Stewardship, learning, and memory in disaster resilience. Environmental Education Research 16(5–6):591–609. https://doi.org/10.1080/13504622.2010.505437 Tran, T. A., and R. Rodela. 2019. Integrating farmers’ adaptive knowledge into flood management and adaptation policies in the Vietnamese Mekong Delta: a social learning perspective. Global Environmental Change 55(May 2018):84–96. https://doi. org/10.1016/j.gloenvcha.2019.02.004 Trenberth, K. 2005. Uncertainty in hurricanes and global warming. Science 308(June):1753–1754. https://doi.org/10.1126/ science.1112551 Trogrlić, R. Š., G. Wright, M. Duncan, M. van den Homberg, A. Adeloye, F. Mwale, and J. Mwafulirwa. 2019. Characterising local knowledge across the flood risk management cycle: a case study of southern Malawi. Sustainability 11(6): 1681. https://doi. org/10.3390/su11061681 Uriarte, M., J. Thompson, and J. K. Zimmerman. 2019. Hurricane María tripled stem breaks and doubled tree mortality relative to other major storms. Nature Communications 10(1):1– 7. https://doi.org/10.1038/s41467-019-09319-2 Wicks, A. J., and D. E. Atkinson. 2017. Identification and classification of storm surge events at Red Dog Dock, Alaska, 2004–2014. Natural Hazards 86(2):877–900. https://doi. org/10.1007/s11069-016-2722-1 Wilson, G. A. 2014. Community resilience: path dependency, lock-in effects and transitional ruptures. Journal of Environmental Planning and Management 57(1):1–26. https://doi. org/10.1080/09640568.2012.741519 Woodruff, J. D., J. L. Irish, and S. J. Camargo. 2013. Coastal flooding by tropical cyclones and sea-level rise. Nature 504 (7478):44–52. https://doi.org/10.1038/nature12855 Zhao, H., and Y. Wang. 2018. Phytoplankton increases induced by tropical cyclones in the South China Sea during 1998–2015. Journal of Geophysical Research: Oceans 123(4):2903–2920. https://doi.org/10.1002/2017JC013549

APPENDIX 1

In scientific literature, Juba and Dala are explained as spring tides (when the earth, the moon, and sun are in alignment). During a new moon, the gravitational forces of the moon and the sun pull along the same direction resulting in high water level (i.e. Juba). During a full moon, the gravitational forces of the sun and the moon exert forces in opposite direction resulting in low level of water (i.e. Dala) (Fig. S.1) (Gönnert and Sossidi, 2011; Park and Suh, 2012). When a high tide coincides with Juba it produces a very high level of tide (i.e. higher than average) and conversely, when low tide coincides with Dala it produces a very low level of tide (i.e. lower than average). During Cyclone Sidr, it was low tide with low spring tide (lower left Quatrain in Fig. S.1).

Fig. S.1: Connection between spring tides with high and low tides

References Gönnert, G., Sossidi, K., 2011. A new approach to calculate extreme storm surges. Irrig. Drain. 60, 91–98. https://doi.org/10.1002/ird.681 Park, Y.H., Suh, K.D., 2012. Variations of storm surge caused by shallow water depths and extreme tidal ranges. Ocean Eng. 55, 44–51. https://doi.org/10.1016/j.oceaneng.2012.07.032